Method and apparatus for processing seismographic data by correlation

Abstract

The invention relates to a method and apparatus for processing data from a seismic survey. In accordance with the invention, the data is regrouped to form a plurality of seismic images (A1, A2, A3, B1, B2, B3, C1, C2, C3). These images being defined by image elements defined in spatial position with respect to an operation direction (D) and having a grey level of predetermined amplitude, the method consisting for each pair of images (Ai,Bj) of carrying out processing by correlation of the amplitude of the grey levels, in order to establish an intercorrelation coefficient and to compensate the localization errors in the position of the images. Application to mono-, bi- or tri-dimensional marine geophysical prospecting.

I claim: 
     
       1. A method of processing data from a seismographic survey by common mid-point gathering in which said data are regrouped to form a plurality of images, wherein said images are defined by image elements defined in spatial position with respect to an operating direction and having a level of predetermined amplitude, said method comprising, for each pair of images, processing by correlation of the amplitude level of said corresponding image elements in order to establish an intercorrelation coefficient for at least one pair of said images in order to compensate for positioning errors of said images. 
     
     
       2. A method according to claim 1, wherein said processing by correlation is a processing by spatial correlation on at least one spatial position dimension of said image elements. 
     
     
       3. A method according to claim 1, wherein said correlation processing is a processing by space time correlation in the time dimension of sad image elements. 
     
     
       4. A method according to claim 1, wherein following said correlation process, said method consists of determining space-time displacements between each pair of images, applying said displacements to values representative of amplitude levels of said image elements considered and summing said corresponding values after displacement for each image pair in order to reduce the signal to noise ratio of the resultant image. 
     
     
       5. A method of three-dimensional processing of marine seismographic survey data by common mid-point gathering, comprising the steps of: obtaining said data from one of a plurality of N acquisition lines permitting definition of n mean lines of substantially equidistant mid points (1 to n) with respect to a moving seismic source, on one operating direction, on each acquisition line and to a plurality of geophones arranged on a prospecting cable associated with each said source;   grouping said data to form a plurality of images (A1, A2, . . . An); (B1, B2 . . . Bn); (C1, C2, . . . Cn) wherein said images (Ai,Bj,Ck) are defined by image elements for an x abscissa in a direction substantially parallel to said operating direction at a given time t, and having a respective amplitude level Ai.sub.(x,t) Bj.sub.(x,t), Ck.sub.(x,t) ;   correlating by amplitude level between corresponding image elements, in order to establish between two distinct images chosen from the group Ai,Bj,Ck an intercorrelation coefficient so as to compensate for data errors due to error in the positioning of said cables.   
     
     
       6. A method according to claim 5, wherein said images Ai, Bj, Ck of which said intercorrelation coefficient is greater than a predetermined value are summed in a manner to reduce the signal to noise ratio. 
     
     
       7. A method according to claim 6, wherein in order to determine position information relating to images Ai, Bj on said direction y transverse to said operating direction, said method consists, for each couple of images Ai, Bj of: establishing a dissimilarity coefficient d ij  satisfying the equation:   d.sub.ij =√1/ρ.sup.2 -Bo     in which ρ represents a value of said spatial intercorrelation coefficient CΔx Ai ,Bj and Bo represents an arbitrary constant value representative of the mean signal to noise ratio for image elements of said images Ai, Bj,     forming from values d ij  of said dissimilarity coefficient, a two-dimensional diagram, said diagram consisting of associating with each image Ai, Bj, and each pair of images Ai,Bj said corresponding coefficients d ij  assimilated to distance between said points Ai and Bj.   
     
     
       8. A method according to claim 5, wherein said correlation processing consists, between two distinct images Ai, Bj of establishing an intercorrelation coefficient of a form: ##EQU3## said correlation processing consisting of a spatial correlation on the x dimension. 
     
     
       9. A method according to claim 8, wherein spatial deviation for minimum correlation (Δx M ) between two images Ai,Bj is obtained for a maximum intercorrelation coefficient CΔx Ai ,Bj. 
     
     
       10. A method according to claim 8, wherein further to a spatial correlation, a time correlation is carried out, said intercorrelation coefficient being of a form: ##EQU4## a space-time deviation (Δx M ,Δt M ) for minimum correlation between two images Ai, Bj being obtained for a maximum intercorrelation coefficient KΔx,Δt Ai ,Bj. 
     
     
       11. A method according to claim 10, wherein said space-time deviation (ΔxM,ΔtM) for any two images Ai, Bj satisfies an equation:   Δt.sub.M =aΔx.sub.M +bΔy.sub.M +cΔ(h.sup.2)+dΔα     in which a,b,c, and d are parameters connected to director cosines of said considered mid point, Δy M  represents deviation of two mid points on said direction y, Δ(h 2 ) represents offset deviation between two images considered, and Δα represents deviation derived from relative prospection cables to acquisition lines Ai, Bj respectively.   
     
     
       12. A method according to claim 10, wherein said method consists further for at least each of said values of spatial Δx M  and/or time Δt M  deviation for correlation, relative to any two images Ai, Bj, of: arranging said intercorrelation coefficients CΔx Ai ,Bj and KΔx,Δt Ai ,Bj in a manner to form for each an array of values,   establishing for each pair of images Ai, Bj a representative deviation diagram for each image Ai, Bj in a space Δt M , Δx M  and mean values of offset for each image.   
     
     
       13. A method according to claim 12, wherein said images of which the values of mean offset are neighbouring, are determined by a position substantially aligned with respect to said dimension Δt M . 
     
     
       14. A device for processing seismographic survey data comprising: computing means for grouping said data to form a plurality of images, said images being defined by image elements, said image elements being defined along a spatial axis, said axis being substantially parallel to an operating direction, said image elements having a level of predetermined amplitude, and for correlating by amplitude level between corresponding image elements in each of said pair of images so as to establish an intercorrelation coefficient to compensate for positioning errors of said images;   permanent memory means coupled to said computing means for storing a set of instructions, said computing means utilizing said instructions to accomplish said grouping and said correlating operations.